Summary

This study aimed to improve the assessment of cognition in HIV-positive individuals by combining traditional pencil-and-paper tests with computerized tasks. The researchers evaluated 75 HIV-positive individuals without dementia using the Montreal Cognitive Assessment (MoCA) and a set of computerized tests. They analyzed the data using Rasch analysis, a statistical method for improving the scoring of measurements based on responses to multiple items, such as several items from the MoCA and performance on computerized tests. The results showed that while the MoCA items could be used to measure cognitive ability, the addition of computerized tests improved the accuracy and precision of the assessment. The results support a relatively straightforward approach to the measurement of cognition as a global construct, in which cognition is evaluated across an appropriate range of difficulty (easy tasks, more difficult tasks) rather than across cognitive abilities (such as memory and concentration). This approach has the potential to quickly and quantitatively detect and monitor cognitive impairment in persons living with HIV.

Abstract

Background

Existing tools for rapid cognitive assessment in HIV-positive individuals with mild cognitive deficits lack sensitivity or do not meet psychometric requirements for tracking changes in cognitive ability over time.

Methods

Seventy-five nondemented HIV-positive patients were evaluated with the Montreal Cognitive Assessment (MoCA), a brief battery of standardized neuropsychological tests, and computerized tasks evaluating frontal-executive function and processing speed. Rasch analyses were applied to the MoCA data set and subsequently to the full set of data from all tests.

Results

The MoCA was found to adequately measure cognitive ability as a single, global construct in this HIV-positive cohort, although it showed poorer precision for measuring patients of higher ability. Combining the additional tests with the MoCA resulted in a battery with better psychometric properties that also better targeted the range of abilities in this cohort.

Conclusion

This application of modern test development techniques shows a path towards a quick, quantitative, global approach to cognitive assessment with promise both for initial detection and for longitudinal follow-up of cognitive impairment in patients with HIV infection.

DOI: 10.1111/j.1468-1293.2010.00910.x

Open Access: PDF of original journal article